|
As my recent series of posts has indicated, I am seeing a lot of future changes in the email industry.
What do I think we can look forward to in email in 2010?
Authentication
In the realm of real authentication, the protocol most are using is is DKIM. While people will probably continue to publish SPF records (and Microsoft will continue to cling to the hope it becomes widespread) its relevance will continue to decrease. As less people pay attention to SPF, records may be unmaintained and become stale further decreasing their use and relevance.
In contrast to SPF, DKIM will continue rolling out. More senders (both the ESPs and the ISPs) will be signing outgoing mail with DKIM. More receivers will be checking DKIM signatures and monitoring domain reputation. I think we’re on the cusp of critical mass and signing will become less of a bonus and more of a given. Right now, it seems that senders who are signing with DKIM are seeing a bit of a reputation bump just because they’re signing. I expect this positive effect will wane, but for now anyone who is signing seems to be seeing improved delivery.
Domain based reputation
Domain based reputation is on the upswing and I see that continuing through 2010. I don’t, however, see domain based reputation replacing or even becoming more important than IP based reputation. A few people have predicted that domain reputation will replace IP reputation, and they’re wrong. Domain based reputation will augment but not replace IP based reputation. It is easy and efficient to check the reputation of a connecting IP address and a receiver can make a preliminary delivery decision without having to accept the full email.
Where domain based reputation will have the biggest effect is for IP addresses with mixed mail streams or IP addresses with no reputation. Small senders often have to share IP addresses with other senders and domain based reputation will allow them to establish their own reputation separately from the reputation of other senders using the same IP. The other real bonus will be when moving mail from one IP to another. Domain based reputation may decrease the time required to warmup an IP address.
Engagement
The buzzword for 2010 is engagement. ISPs will be measuring engagement and making delivery decisions based on how much their users want particular email. In the past ISPs have used measurements like complaint rates and bounce rates to measure how wanted email is. These numbers correlate with how wanted mail is, but are relatively easy for senders to game. In 2010, ISPs are going to actually start filtering based on how wanted mail is. “Wanted” mail will no longer be measured using the proxy measurements, as those have proven to be easy to game. Instead, ISPs will directly measure how much recipients want a particular mail. These changes will force senders to stop sending mail that does not generate complaints and start sending mails that recipients are eager to receive.
Social Networking
I don’t see social networking replacing email marketing at any time. I do see, though, email marketing giving recipients opportunities to share information with social networks. Smart senders will provide easy links so that recipients can share information with their social networks. When marketers do well, they’ll have happy recipients who want to share the information. When marketers do poorly, however, they will have to deal with unhappy recipients. It only takes a few people publicizing a company failure to generate negative buzz.
Conclusion
In 2010 email marketing is going to get much more challenging for everyone. Recipients, and their ISPs, are expecting more and better things from email marketing. Senders who are currently meeting expectations may struggle to meet those increasing standards within their current marketing frameworks. Successful marketers will be able to make the switch from sending mail that doesn’t annoy customers to sending mail that recipients truly want. On the ESP side, they may find they had to provide more guidance and consulting support for customers. They may also need to change some policies and improve their problem detection systems.
This is the year of engagement, and senders can’t fake engagement the way they can other metrics. Marginal senders will struggle to adapt to the new conditions. Better senders will need to change some things, but will improve their marketing to meet the new standards. Overall, though, the changes will drive all senders to really send mail people want. This leads to more engaged recipients. More engaged recipients leads to better delivery and better ROI for those marketers as well as a better inbox experience for recipients.
Sponsored byVerisign
Sponsored byVerisign
Sponsored byRadix
Sponsored byIPv4.Global
Sponsored byDNIB.com
Sponsored byCSC
Sponsored byWhoisXML API
Someone already said that spam filtering engines have moved into the world of AI. However, distinguishing mail that doesn’t annoy customers from mail that recipients truly want, apparently implies a level of psychological insight that is normally found in “natural” intelligence only. It is not clear to me what kind of direct interaction between an ISP and its users will yield that measurement.
Furthermore, I’m puzzled about the usefulness of that kind of distinction for shaping mail protocols. A solicitation of payment may be an example of a legitimate message that someone may not truly want to receive: should it be delivered, then?
As for authentication and reputation, I’m confused by the seemingly contradictory stance toward SPF and IP based reputation, as they are both characterized by the same ease and efficiency of operations. On the other hand, DKIM requires an Identity Assessor that has to be based on mail domain reputation. The difference between those reputation systems is much more than a mere change of identifier: A domain reputation service worth its salt should be able to verify how a given domain’s admins banish spammers. (In turn, the latter implies that a well reputed mail domain should know—not necessarily disclose—the identities of its users.) I haven’t seen one, yet.